CN103346983A - OFDM self-adaption complex interpolation channel estimation method based on comb-type pilot frequency - Google Patents

Abstract

The invention discloses an OFDM self-adaption complex interpolation channel estimation method based on comb-type pilot frequency. A set of maximum multipath time delay values arranged from large to small is confirmed, corresponding complex interpolation filters are designed according to the complex interpolation algorithm, users try to match complex interpolation filter coefficients with channel impact response one by one until the channel energy ratio between average channel energy and standard channel energy is not more than a threshold value of the channel energy ratio for the first time, the complex interpolation filter coefficient corresponding to the previous one maximum multipath time delay value is the required complex interpolation filter coefficient, and therefore self-adaption confirmation of the complex interpolation filter coefficients is achieved. According to the OFDM self-adaption complex interpolation channel estimation method, the complex interpolation algorithm is adopted, and longer multipath time delay can be resisted; the complex interpolation filter coefficients are adopted for being matched with the channel impact response, and influences on channel estimation caused by noise and channel impact response mirror components can be reduced; time-varying channels can be tracked by updating the channel energy ratio in real time, and the complex interpolation filter coefficients are more accurate.

Description

An OFDM Adaptive Complex Interpolation Channel Estimation Method Based on Comb Pilots

technical field

The invention belongs to the technical field of OFDM communication, and more specifically, relates to an OFDM adaptive complex interpolation channel estimation method based on comb pilots. the

Background technique

OFDM (Orthogonal Frequency Division Multiplexing, Orthogonal Frequency Division Multiplexing) is a multi-carrier modulation technology that has natural advantages in combating multipath fading and is very suitable for high-speed data transmission. Therefore, OFDM is used in modern wireless broadband access systems. It has been widely used, such as DAB (Digital Audio Broadcasting, Digital Audio Broadcasting), DVB (Digital Video Broadcasting, Digital Video Broadcasting), 3G, LTE (Long Term Evolution, Long Term Evolution), Wi-Fi, WiMax, etc. In a wireless OFDM system, the multipath effect and the Doppler effect will cause the wireless channel to have frequency-domain selective fading and time-selective fading respectively, which will have a bad impact on the receiver using coherent demodulation and degrade the system performance. Therefore, a high-performance channel estimation method is required to accurately obtain channel information and eliminate the influence of multipath channels through channel equalization. Designing a low-complexity channel estimator to quickly track changing channel information is a key technology for designing high-performance OFDM receivers. There are many channel estimation algorithms, and the most commonly used channel estimation algorithms are LS (Least Square, least squares) channel estimation and MMSE (Minimum Mean-Square Error, minimum mean square error) channel estimation. LS channel estimation is simple but susceptible to noise; MMSE channel estimation has smaller errors than LS estimation under the same conditions, but it needs the autocorrelation and noise statistics of the channel, so it is more complicated. the

Single frequency network (SFN: Single Frequency Network) has been widely used in recent years because of its networking mode: high spectrum utilization, can expand effective coverage, suitable for mobile, portable reception, etc., has been widely used in recent years, such as the European DVB- T/H standard etc. In a single frequency network, all transmitters transmit the same data on the same frequency at the same time, so when the receiver receives data from transmitters in adjacent cells or farther cells, it will inevitably generate long delay and multipath delay. Long-latency multipath delays can cause severe frequency-selective fading, which cannot be handled in single-carrier systems. Because for the same bandwidth, the symbol period of the single-carrier system is much shorter than that of the multi-carrier system, so it is much more sensitive to frequency selective fading. In multi-carrier systems, traditional channel estimation algorithms also have shortcomings in dealing with SFN long-delay channels. the

In the OFDM system, a pilot-based channel estimation algorithm is often used, that is, the CFR (Channel Frequency Response, channel frequency domain response) at the pilot is first estimated, and then the CFR at the data is estimated by an interpolation algorithm. The pilot frequency is a group of random sequences, and is known to the receiving end, so the CFR at the pilot frequency can be estimated by receiving the signal through the pilot frequency. The channel estimation at the pilot frequency can be performed by using the LS algorithm and the MMSE algorithm. the

In terms of interpolation algorithms, traditional channel estimation algorithms generally use real coefficient interpolation algorithms, including polynomial interpolation and digital interpolation filter interpolation. Polynomial interpolation also includes linear interpolation, second-order Gaussian interpolation, cubic Lagrangian interpolation, cubic spline interpolation, etc. Digital interpolation filter interpolation includes low-pass raised cosine interpolation and the like. According to the properties of Fourier transform, the time-domain window obtained by Fourier transform of real interpolation coefficients is symmetrical. Combined with the Nyquist sampling theorem, the maximum multipath delay τ _max must satisfy τ _max <T _u /2D, Where T _u is the OFDM symbol period, and D is the frequency domain pilot interval. This requirement may not be satisfied under the SFN long-delay channel. Taking the SFN channel in the DVB-T system as an example, τ _max <T _u /2D≈150us, but the actual longest multipath delay may exceed 200μs. In this case, the real coefficient interpolation algorithm cannot meet the requirements, so the industry proposes a complex interpolation algorithm to process the long-delay SFN channel. Figure 1 is a comparison diagram of the ideal real interpolation and complex interpolation low-pass filters against multipath delay. As shown in Figure 1, the SFN channel using the complex interpolation algorithm can resist greater multipath delay. However, the current complex interpolation algorithm uses fixed complex interpolation coefficients and cannot track time-varying channels. Therefore, when the CIR (Channel Impulse Response, channel impulse response) does not match the time domain window obtained by Fourier transform of the complex interpolation coefficients, Will cause system performance to degrade.

Contents of the invention

The purpose of the present invention is to overcome the deficiencies of the prior art, to provide a comb pilot-based OFDM adaptive complex interpolation channel estimation method, using multiple complex interpolation filter coefficients to match the CIR to optimize the complex interpolation filter coefficients , so as to track the time-varying channel in real time and make the channel estimation more accurate. the

In order to realize the foregoing invention object, the present invention is based on the OFDM adaptive complex interpolation channel estimation method of comb pilot, it is characterized in that comprising:

(1) Estimate the channel frequency domain response at the comb pilot, and obtain the channel frequency domain response at the comb pilot after zero interpolation

(2) Construct an integer set Ψ={τ ₀ , τ ₁ , ..., τ _J-1 }, the number of set elements J is set according to the actual situation, and satisfies τ ₀ ＞τ ₁ ＞…＞τ _J-1 ＞0 , the maximum value τ ₀ =N _g , N _g represents the number of subcarriers included in the OFDM cyclic prefix;

(3) Use the widest window (MWW: Maximum Width Window) method to calculate the reference channel energy E[τ ₀ ] corresponding to τ ₀ , and set the one-dimensional frequency domain interpolation (FDI, Frequency Domain Interpolation) output of the comb pilot The channel energy ratio threshold value T _h of the average channel energy and the reference channel energy;

对应的复插值滤波器系数

即是所需的复插值滤波器系数；平均信道能量E[τ_i]的计算方法为：  (4) Calculate the corresponding average channel energy E[τ i ] of τ _i , 1≤i≤J-1 in turn, once E[τ _{i ]} /E[τ ₀ ]≤T _h , then

Corresponding complex interpolation filter coefficients

That is, the required complex interpolation filter coefficient; the calculation method of the average channel energy E[τ _i ] is:

n＝-Q,…,Q，Q＝(N_q-1)/2，其中N_q为复插值滤波器阶数，根据实际情况设置；  4.1) Make the maximum multipath delay τ _max = τ _i T _s , T _s represents the sampling time of the OFDM symbol, and use the complex interpolation algorithm to calculate the complex interpolation filter coefficients

n=-Q,...,Q, Q=(N _q -1)/2, where N _q is the order of the complex interpolation filter, set according to the actual situation;

4.2), calculate the FDI output corresponding to τ _i :

${\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; h</span>}}}^{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; ]</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; Q</span>}}{\overset{\mathrm{<span\; class="notranslate">\; Q</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; b</span>}}^{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ]</span>\stackrel{<span\; class="notranslate">\; ~</span>}{\mathrm{<span\; class="notranslate">\; h</span>}}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ]</span><span\; class="notranslate">\; ;</span>$

4.3), calculate the average channel energy E[τ _i ] corresponding to τ _i :

$\mathrm{<span\; class="notranslate">\; E.</span>}<span\; class="notranslate">\; [</span>{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; ]</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; N</span>}}\underset{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}}{\overset{\mathrm{<span\; class="notranslate">\; N</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{<span\; class="notranslate">\; |</span>{\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; h</span>}}}^{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; ]</span><span\; class="notranslate">\; |</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}$

Among them, N is the number of effective subcarriers of OFDM symbols;

(5), according to the complex interpolation filter coefficient obtained in step (4) Compute the frequency-domain response of the channel at the data: ${\hat{h}}^{\left(\widehat{\mathrm{\&tau;}}\right)}\left[k\right]=\underset{\mathrm{no}=-Q}{\overset{Q}{\mathrm{\&Sigma;}}}{b}^{\left(\widehat{\mathrm{\&tau;}}\right)}\left[\mathrm{no}\right]\tilde{h}[k-\mathrm{no}].$

Among them, the complex interpolation algorithm includes the following steps:

4.1.1), determine the complex interpolation filter parameters:

其中Δf为OFDM符号的子载波间隔；  passband frequency

Where Δf is the subcarrier spacing of the OFDM symbol;

其中D为频域导频间隔；  Transition Bandwidth

where D is the frequency domain pilot interval;

Cut-off frequency $t_e=t_p+\frac{1}{2}{t}_{\mathrm{sp}};$

4.1.2) Calculate the weighted window function

4.1.3), calculate real interpolation filter coefficients

4.1.4), calculate complex interpolation filter coefficients

Further, the OFDM adaptive complex interpolation channel estimation method also includes steps:

对应的信道能量比

如果

返回步骤（2）重新进行复插值滤波器系数计算与信道估计，否则复插值滤波器系数保持不变。  (6). Set the update cycle of adaptive complex interpolation. Whenever the cycle time arrives, update

Corresponding channel energy ratio

if

Return to step (2) to recalculate the complex interpolation filter coefficients and channel estimation, otherwise the complex interpolation filter coefficients remain unchanged.

The OFDM adaptive complex interpolation channel estimation method based on comb pilots in the present invention adopts a tentative mechanism, firstly determine a set of maximum multipath delay values arranged from large to small, and then design the corresponding complex interpolation algorithm according to the complex interpolation algorithm. Interpolation filter, and then use the time domain windows obtained by these complex interpolation filter coefficients to try to match the CIR one by one, until the channel energy ratio of the average channel energy output by FDI relative to the reference channel energy obtained by the widest window method is not greater than for the first time The channel energy ratio threshold T _h and the complex interpolation filter coefficient corresponding to the last multipath time delay are the required complex interpolation filter coefficients, so as to realize the adaptive determination of the complex interpolation filter coefficients.

The OFDM adaptive complex interpolation channel estimation method based on the comb pilot in the present invention can achieve the following beneficial effects:

①. Compared with the traditional real coefficient interpolation algorithm, the present invention can resist greater multipath delay due to the use of complex interpolation algorithm;

②, the present invention matches CIR by adopting a plurality of complex interpolation filter coefficients, thus obtained complex interpolation filter coefficients can reduce the impact of noise and CIR image components on channel estimation;

③. The present invention can also track the time-varying channel by updating the channel energy ratio in real time, so that the coefficients of the complex interpolation filter can be adaptively changed, thereby making the channel estimation more accurate;

可以事先计算好，因此可以快速跟踪信道变化。  ④. For a certain set of integers Ψ,

Can be calculated in advance, so channel changes can be tracked quickly.

Description of drawings

Figure 1 is a comparison diagram of ideal real interpolation and complex interpolation low-pass filter against multipath time delay;

Fig. 2 is a schematic diagram of OFDM system adopting the present invention;

Fig. 3 is the conversion schematic diagram of two-dimensional scattered pilot;

Fig. 4 is a kind of specific embodiment flowchart of the OFDM adaptive multiple interpolation channel estimation method based on comb pilot of the present invention;

Fig. 5 is a schematic diagram of real interpolation filter time-domain parameters;

Figure 6 is a schematic diagram of the influence of SFN echo delay on bit error performance;

Figure 7 is a bit error performance simulation under the AWGN channel;

Figure 8 is the simulation of bit error performance under the Nokia handheld channel. the

Detailed ways

Specific embodiments of the present invention will be described below in conjunction with the accompanying drawings, so that those skilled in the art can better understand the present invention. It should be noted that in the following description, when detailed descriptions of known functions and designs may dilute the main content of the present invention, these descriptions will be omitted here. the

Example

Fig. 2 is a schematic diagram of an OFDM system adopting the present invention. As shown in Figure 2, adopt the OFDM system of the present invention, obtain the CFR at the pilot frequency through the comb pilot at the receiving end, then carry out adaptive adjustment to the complex interpolation filter coefficient according to the present invention, according to the obtained complex interpolation filter coefficients for channel estimation. The present invention aims at the OFDM system in which the pilot frequency structure adopts the comb pilot, and the comb pilot only performs FDI. The two-dimensional scattered pilot can obtain the estimated value of the virtual pilot by interpolating in the time direction (TDI: Time Direction Interpolation). Fig. 3 is a schematic diagram of conversion of two-dimensional scattered pilots. As shown in FIG. 3 , the two-dimensional scattered pilots are transformed into comb pilots through TDI, so the present invention is also applicable to OFDM systems in which the pilot structure adopts two-dimensional scattered pilots. the

Fig. 4 is a flow chart of a specific embodiment of the comb pilot-based OFDM adaptive multiple interpolation channel estimation method of the present invention. As shown in Figure 4, the present invention comprises the following steps:

S401: Channel estimation at the pilot, that is, estimating the channel frequency domain response at the comb pilot, and obtaining the channel frequency domain response at the comb pilot after zero interpolation

In this embodiment, the channel estimation at the comb-shaped pilot uses the LS algorithm to obtain the frequency response at the comb-shaped pilot after zero interpolation:

Among them, X[k] is the pilot transmission signal known by the receiving end, Y[k] is the pilot receiving signal received by the receiving end, and k is the frequency point of the pilot in the frequency domain. the

S402: Construct an integer set Ψ={τ ₀ , τ ₁ , ..., τ _J-1 }, the number of set elements J is set according to the actual situation, and satisfies τ ₀ >τ ₁ >... >τ _J-1 >0, the maximum Value τ ₀ =N _g , where N _g represents the number of subcarriers included in the OFDM cyclic prefix.

S403: Using the MWW method to calculate the reference channel energy E[τ ₀ ] corresponding to τ ₀ , and setting the threshold value T _h of the channel energy ratio between the average channel energy output by the one-dimensional frequency domain interpolation FDI of the comb pilot and the reference channel energy.

The MWW method is a method in which the maximum multipath delay τ _max takes the length of the cyclic prefix (CP: Cyclic Prefix) for complex interpolation design, that is, τ _max = N _g T _s , where: N _g represents the number of subcarriers included in the CP, and T _s represents the sampling time of the OFDM symbol. The channel energy obtained by using the MWW method is the largest, so the present invention uses this as the reference channel energy. As the width of the window function we choose decreases, the resulting energy decreases. The criterion for selecting the width of the window function is: the width of the window should be as small as possible, so that more noise can be filtered out, but the width of the window should not be too small, so as not to filter out the required CIR part. Therefore, in practical applications, the channel energy is usually higher than the gate The limit T _h is chosen to be slightly smaller than 1. The energy ratio threshold Th _h is selected according to two channel parameters, power delay spectrum and signal-to-noise ratio. In practical applications, a suitable value can be obtained through a large number of experiments for the channel.

S404: Initialize i=1;

S405: Calculate the FDI output result corresponding to τ _i , the specific calculation method is:

n＝-Q,…,Q，Q为复插值滤波器长度，Q＝(N_q-1)/2，其中N_q为复插值滤波器阶数，根据实际情况设置。在实际应用中，N_q的取值需要在复杂度和性能之间取折中，可通过尝试方法获得。  5.1) Make the maximum multipath delay τ _max = τ _i T _s , where T _s represents the sampling time of the OFDM symbol, and use the complex interpolation algorithm to calculate the complex interpolation filter coefficients

n=-Q,...,Q, Q is the length of the complex interpolation filter, Q=(N _q -1)/2, where N _q is the order of the complex interpolation filter, which is set according to the actual situation. In practical applications, the value of N _q needs to be a compromise between complexity and performance, which can be obtained by trial and error.

The complex interpolation algorithm usually includes two steps: the first step is to design the corresponding real interpolation filter; the second step is to convert the real interpolation filter into the required complex interpolation filter. There are many design methods for real interpolation filters, such as FIR (Finite Impulse Response, finite impulse response), IIR (Infinite Impulse Response, infinite impulse response), Parks-McClellan, etc. Among them, the most widely used is the FIR design method of the low-pass sinc windowing function. There are many options for window functions, such as rectangular window, triangular window, Hanning window, Hamming window, Kaiser window, etc. the

Fig. 5 is a schematic diagram of time-domain parameters of a real interpolation filter. In the present embodiment, take the low-pass sinc plus Kaiser window as an example, and illustrate the steps of complex interpolation filter coefficient calculation in conjunction with Figure 5, including:

5.1.1), determine the real interpolation filter time domain parameters, including:

其中Δf为OFDM符号的子载波间隔；  passband frequency

Where Δf is the subcarrier spacing of the OFDM symbol;

Transition Bandwidth Among them, D is the comb-shaped pilot interval in the frequency domain.

Cut-off frequency $t_e=t_p+\frac{1}{2}{t}_{\mathrm{sp}}.$

5.1.2) Calculate the weighted window function

其中I₀(·)表示第一类零阶修正贝塞尔函数，α为Kaiser窗的形状因子：  What adopted in this embodiment is Kaiser window, and its window function

Among them, I ₀ (·) represents the first kind of zero-order modified Bessel function, and α is the shape factor of the Kaiser window:

$\mathrm{<span\; class="notranslate">\; \&alpha;</span>}<span\; class="notranslate">\; =</span>\left\{\begin{array}{c}\mathrm{<span\; class="notranslate">\; 0.1102</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; A</span>}}_{\mathrm{<span\; class="notranslate">\; the\; s</span>}}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 8.7</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; A</span>}}_{\mathrm{<span\; class="notranslate">\; the\; s</span>}}<span\; class="notranslate">\; \&Greater\; Equal;</span>\mathrm{<span\; class="notranslate">\; 50</span>}\\ \mathrm{<span\; class="notranslate">\; 0.5842</span>}{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; A</span>}}_{\mathrm{<span\; class="notranslate">\; the\; s</span>}}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; twenty\; one</span>}<span\; class="notranslate">\; )</span>}^{\mathrm{<span\; class="notranslate">\; 0.4</span>}}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 0.07886</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; A</span>}}_{\mathrm{<span\; class="notranslate">\; the\; s</span>}}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; twenty\; one</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; twenty\; one</span>}<span\; class="notranslate">\; <</span>{\mathrm{<span\; class="notranslate">\; SA</span>}}_{\mathrm{<span\; class="notranslate">\; the\; s</span>}}<span\; class="notranslate">\; <</span>\mathrm{<span\; class="notranslate">\; 50</span>}\end{array}\right.<span\; class="notranslate">\; .</span>$

A _s =14.36(N _q -1)t _sp +7.95, which is the stopband attenuation of the real interpolation filter.

5.1.3), calculate real interpolation filter coefficients

5.1.4), calculate complex interpolation filter coefficients

5.2), calculate the FDI output result of the comb pilot:

${\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; h</span>}}}^{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; ]</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; =</span><span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; Q</span>}}{\overset{\mathrm{<span\; class="notranslate">\; Q</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; b</span>}}^{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ]</span>\stackrel{<span\; class="notranslate">\; ~</span>}{\mathrm{<span\; class="notranslate">\; h</span>}}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ]</span><span\; class="notranslate">\; .</span>$

经傅里叶变换可得到时域窗，

经博立叶变换可得到CIR，本步骤即实现了时域窗与CIR的匹配。  complex interpolation filter coefficients

The time domain window can be obtained by Fourier transform,

The CIR can be obtained through the Boyier transform, and this step realizes the matching between the time domain window and the CIR.

S406: Calculate the average channel energy E[τ _i ] corresponding to τ _i :

$\mathrm{<span\; class="notranslate">\; E.</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; N</span>}}\underset{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}}{\overset{\mathrm{<span\; class="notranslate">\; N</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{<span\; class="notranslate">\; |</span>{\stackrel{<span\; class="notranslate">\; ^</span>}{\mathrm{<span\; class="notranslate">\; h</span>}}}^{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&tau;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; [</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; ]</span><span\; class="notranslate">\; |</span>}^{\mathrm{<span\; class="notranslate">\; 2</span>}}$

Wherein, N is the effective number of subcarriers of the OFDM symbol. the

则

对应的复插值滤波器系数

即是所需的复插值滤波器系数，进入步骤S409；如果E[τ_i]/E[τ₀]＞T_h，进入步骤S408。  S407: Judging whether E[τ _i ]/E[τ ₀ ]≤T _h , if E[τ _i ]/E[τ ₀ ]≤T _h , let

but

Corresponding complex interpolation filter coefficients

That is, the required complex interpolation filter coefficient, go to step S409; if E[τ _i ]/E[τ ₀ ]>T _h , go to step S408.

S408: i=i+1, return to step S405. the

S409: Output the obtained complex interpolation filter coefficients Go to step S410.

S410: Calculate the frequency domain response of the channel at the data point:

经FDI输出的平均信道能量与MWW方法得到的信道能量均会随着时间变化，信道能量比E[τ_i]/E[τ₀]也会随着时间变化。为了实现对时变信道的跟踪，在实际应用中，设置自适应复插值更新周期，每当周期时间到来时，更新

对应的信道能量比E[τ_i]/E[τ₀]，如果E[τ_i]/E[τ₀]＞T_h，返回步骤S402重新进行复插值滤波器系数计算，否则复插值滤波器系数保持不变。  Under the time-varying channel, the complex interpolation filter coefficient obtained above

Both the average channel energy output by FDI and the channel energy obtained by MWW method will change with time, and the channel energy ratio E[τ _i ]/E[τ ₀ ] will also change with time. In order to realize the tracking of the time-varying channel, in practical applications, the adaptive complex interpolation update cycle is set, and whenever the cycle time comes, update

Corresponding channel energy ratio E[τ _i ]/E[τ ₀ ], if E[τ _i ]/E[τ ₀ ]>T _h , return to step S402 to recalculate the coefficients of the complex interpolation filter, otherwise the complex interpolation filter Coefficients remain the same.

Compared with the traditional real-coefficient interpolation algorithm, the present invention can resist greater multipath time delay due to the adoption of complex interpolation algorithm, and is especially suitable for SFN systems. The present invention matches CIR by adopting multiple complex interpolation filters, and the obtained complex interpolation filter coefficients can reduce the influence of noise and CIR image components on channel estimation. The present invention can also track the time-varying channel by updating the channel energy ratio, so that the complex interpolation filter coefficient can be changed adaptively, so that the channel estimation is more accurate. At the same time, because for a certain set of integers Ψ, Can be calculated in advance, so channel changes can be tracked quickly.

A specific implementation case in the DVB-T system is simulated by using the comb pilot-based OFDM adaptive multiple interpolation channel estimation method described in this embodiment. The simulation parameters are: the number of FFT points is 8192, the constellation mode is 64QAM, the OFDM symbol period is 896μs, the CP mode is 1/4, the frequency domain pilot interval D=3, the filter order N _q =25, and the simulation system adopts Convolutional coding with code rate 2/3. Channel parameters: long delay two-path SFN channel, AWGN channel and short delay Nokia handheld channel. Among them: the two-path SFN channel contains a 0dB echo channel. Table 1 is the power delay spectrum of the Nokia handheld channel.

Table 1

Set the integer set Ψ={2048,1024,512,256,128,64}, in the three channels, the channel energy ratio T _h is 0.91.

In the simulation result graph, since the performance loss introduced by the MMSE algorithm is almost negligible, its performance is taken as a reference standard in this embodiment. Fig. 6 is a schematic diagram of the influence of the SFN echo delay on the bit error performance. As shown in Figure 6, the signal-to-noise ratio SNR is set to 22dB, and the order N _q of the low-pass raised cosine interpolator in the figure is 49. It can be seen from Figure 6 that the higher the order of polynomial interpolation, the longer the maximum multipath delay that can be resisted. And low-pass raised cosine interpolation performs better than polynomial interpolation. The performance of the self-adaptive complex interpolation algorithm is much better than the first two, and is similar to the MMSE algorithm. Fig. 7 is the bit error performance simulation under the AWGN channel. Figure 8 is the simulation of bit error performance under the Nokia handheld channel. It can be seen from Fig. 7 and Fig. 8 that under these two channels, the performance of the real coefficient interpolation method is similar to that of the MWW method, while the performance of the adaptive complex interpolation algorithm is better than both of them, which is similar to that of the MMSE algorithm.

Although the illustrative specific embodiments of the present invention have been described above, so that those skilled in the art can understand the present invention, it should be clear that the present invention is not limited to the scope of the specific embodiments. For those of ordinary skill in the art, As long as various changes are within the spirit and scope of the present invention defined and determined by the appended claims, these changes are obvious, and all inventions and creations using the concept of the present invention are included in the protection list. the

Claims (3)

1. the OFDM self adaptation based on Comb Pilot is answered the interpolation channel estimation methods, it is characterized in that may further comprise the steps:

(1), estimate the channel frequency domain response at Comb Pilot place, obtain the channel frequency domain response at the Comb Pilot place behind the zero insertion

(2), an integer set of structure Ψ={ τ ₀, τ ₁..., τ _J-1, set element quantity J is according to the actual conditions setting, and satisfies τ ₀＞τ ₁＞...＞τ _J-1＞0, maximum τ ₀=N _g, N _gThe subcarrier number that expression OFDM Cyclic Prefix comprises;

(3), adopt the wideest window method to calculate τ ₀Corresponding reference channel energy E [τ ₀], the channel energy of average channel energy and reference channel energy that the one dimension frequency domain interpolation FDI output of Comb Pilot is set compares threshold T _h

(4), calculate τ successively _i, the average channel energy E [τ of the correspondence of 1≤i≤J-1 _i], in case E[τ _i]/E[τ ₀]≤T _h, then

Corresponding multiple interpolation filter coefficient It namely is required multiple interpolation filter coefficient; Average channel energy E [τ _i] computational methods be:

4.1), make maximum multipath time delay τ _Max=τ _iT _s, T _sIn the sample time of expression OFDM symbol, adopt multiple interpolation algorithm to calculate multiple interpolation filter coefficient

N=-Q ..., Q, Q=(N _q-1)/2, N wherein _qFor multiple interpolation filter exponent number, according to the actual conditions setting;

4.2), calculate τ _iCorresponding FDI output result:

${\hat{H}}^{\left({\mathrm{\&tau;}}_i\right)}\left[k\right]=\underset{n=-Q}{\overset{Q}{\mathrm{\&Sigma;}}}{b}^{\left({\mathrm{\&tau;}}_i\right)}\left[n\right]\tilde{H}[k-n];$

4.3), calculate τ _iCorresponding average channel energy E [τ _i]:

$E\left[{\mathrm{\&tau;}}_i\right]=\frac{1}{N}\underset{k=0}{\overset{N-1}{\mathrm{\&Sigma;}}}{\left|{\hat{H}}^{\left({\mathrm{\&tau;}}_i\right)}\right[k\left]\right|}^2$

Wherein, N is effective subcarrier number of OFDM symbol;

(5), according to the multiple interpolation filter coefficient that obtains in the step (4) The channel frequency domain response at calculated data place: ${\hat{H}}^{\left(\widehat{\mathrm{\&tau;}}\right)}\left[k\right]=\underset{n=-Q}{\overset{Q}{\mathrm{\&Sigma;}}}{b}^{\left(\widehat{\mathrm{\&tau;}}\right)}\left[n\right]\tilde{H}[k-n].$

2. the multiple interpolation channel estimation methods of OFDM self adaptation according to claim 1 is characterized in that, described multiple interpolation algorithm may further comprise the steps:

4.1.1), determine multiple interpolation filter parameter:

Band connection frequency

Wherein Δ f is the subcarrier spacing of OFDM symbol;

The transition band bandwidth

Wherein D is the pilot tone interval;

Cut-off frequency $t_e=t_p+\frac{1}{2}{t}_{\mathrm{sp}};$

4.1.2) calculating weighting windows function

4.1.3), calculate real interpolation filter coefficient

4.1.4), calculate multiple interpolation filter coefficient

3. the multiple interpolation channel estimation methods of OFDM self adaptation according to claim 1 and 2 is characterized in that, also comprises step:

(6), the multiple interpolation update cycle of self adaptation is set, when arrive cycle time, upgrade

Corresponding channel energy ratio

If

Return step (2) and carry out multiple interpolation filter coefficient calculations and channel estimating again, otherwise multiple interpolation filter coefficient remains unchanged.

Images